The homogeneous 3 alpha-hydroxysteroid dehydrogenase (3 alpha-HSD) of rat liver cytosol is potently inhibited at its active site by nonsteroidal and steroidal antiinflammatory drugs (NSAIDS and SAIDS). It binds prostaglandins with high affinity and may function as a 9-hydroxyprostaglandin dehydrogenase. The enzyme fulfills many of the criteria attributed to a target for antiinflammatory drugs. It has been purified in milligram amounts, crystallized, and its cDNA has been cloned. Since the binding of antiinflammatory drugs at the active site predicts their pharmacological potency, the topography of this site will be elucidated using a multi-faceted approach. To solve the 3-dimensional structure of the enzyme, new crystal forms will be grown which diffract beyond the 6 A degree level. A near full length cDNA for 3 alpha-HSD will be sub-cloned into the sequencing vector pGEM-3 and its sequence determined by dideoxy- sequencing. Knowledge of the primary structure will compliment the mapping of catalytic domains of the enzyme. This will be achieved using a series of suicide substrates based on either 4- nitrobenzaldehyde or Nphenylanthranilic acids. The former series labels the NAD(P)+ binding site and the latter series mimics NSAIDS, and have the potential to alkylate the antiinflammatory drug binding site. By using radiolabeled suicide substrates and peptide mapping procedures, peptides residing at each domain will be identified. Methods which permit the placement of bromoacetylating agents in a positional manner around NSAIDS (indomethacin and meclofenamate) and SAIDS are being developed. These agents will be used to identify amino acids that play a functional importance in the binding of antiinflammatory drugs. To assist in the orientation of prostaglandins at the active site, the 9, 11-and 15-hydroxyprostaglandin dehydrogenase activities will be further characterized. Computer (PROPHET) modeling of bound ligands (NSAIDS, SAIDS and prostaglandins) will lead to the generation of an enzyme excluded volume map, i.e., that volume not occupied by the protein but essential for ligand binding. The relationship, that exists between amino acids identified in the affinity labeling studies and bound ligands, will be revealed by placing these amino acids around the perimeter of this volume map. These studies may provide information for a rational approach to the design of superior antiinflammatory drugs.